Preamble

Processing

Def functions and variables

Base

Pairs

Globals

(legacy) Import and process data

Since removing from the server it was running on with complptete access to live calcualtions as they're coming in, the data has been exported to csv files to be imported cleanly.

I could have converted these blocks to raw, but then my code hiding extension wouldn't hide them.

Base

Pairs

Export data to csv for future easy import

(new) Import and process data

Lone Ion Plots

HOMO and $\Delta$HOMO

Please disregard Benzene-Ni

SAPT Energies

Excluding Benzene-Ni

Correlation Plot

Coloured by Ion Name

Excluding Pd$^{2+}$, Ni$^{2+}$ and hard ions(Cl$^-$, Na$^+$ and Li$^+$)

dplusel is dispersion + electrostatics

Coloured by Ion Class (Anion/Cation)

Excluding Pd$^{2+}$, Ni$^{2+}$ and hard ions(Cl$^-$, Na$^+$ and Li$^+$)

dplusel is dispersion + electrostatics

Ion Pair Plots

HOMO and $\Delta$HOMO

SAPT Energy

Correlation Plot

Both

Correlation Plot

Ion pairs and Individual ions combined

Excluding Pd$^{2+}$, Ni$^{2+}$, and hard ions(Cl$^-$, Na$^+$ and Li$^+$)

Katya's Suggestions

1) Correlations

Except Li (Na is missing for some reason) the electrostatics follows the total interaction energy. Could you please extract the gradient from the electrostatics-total graph. I believe that Induction + dispersion + exchange might be summing to within a few kJ mol-1.... Ultimately, electrostatics is the force that drives this interaction... Li definitely makes the bond covalent and hence such a huge difference between electrostatics and total energy.

Total - Electrostatics

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

Total - Dispersion + Electrostatics

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

Total - Dispersion + Electrostatics + Exchange

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

2) Further Correlations

Electrostatics and exchange also correlate with each other, and so do a) dispersion and exchange and b) induction and electrostatics. Extracting their gradients would be good too. Surprisingly, dispersion correlates with exchange without the need to add electrostatics! Although the latter does improve the correlation.... The dispersion-exchange correlation could be compared to Anh's data.... I think this is the reason why her DFT works so well...

Electrostatics - Exchange

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

Dispersion - Exchange

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

Induction - Electrostatics

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$

Excluding Ni$^{2+}$, Pd$^{2+}$, Li$^+$, Na$^+$, Cl$^+$

3) Correlation of ONLY ionpairs and their corresponding lone counterparts

I would like to know where the ion pairs are located on 3.1.1 plots. Could you please include only those of the lowest energy? Would you be able to make SAPT energy graphs combining the ion pairs and the results for individual ions included in these ion pairs?

6) Induction - Electrostatics SAPT comparison

Ni and Pd form covalent bonds by the look of it. In both cases it is the induction that dominates the interaction (hence, the symmetry is crucial). Electrostatics is stronger in Ni but is compensated by high exchange... So induction is the key whereas in ILs pure electrostatics is the key! It must come down to the orbital symmetry I guess. This is very cool though as clearly transition metals favour certain sites whereas the problem of ions is that they lead to non-directional interaction... So, the next goal would be to identify ions that have larger induction over electrostatics (if it is even possible).

This problem seems to be dependent on the molecule with which the ion is interacting and unfortunately we can only really say that BF4$^+$ and Li$^-$ have the highest number of incidence where induction is greater than electrostatics.